Support for JPG Files in Zerox

[shawn8888]

I appreciate the work done on the Zerox project for converting PDF documents to Markdown. However, I've encountered a limitation that I believe could enhance the usability of the tool.

Sometimes, I come across documents in JPG format, and it can be inconvenient to convert them to PDF just to use Zerox for conversion. Would it be possible to add support for directly processing .jpg files? This feature would streamline the workflow and make Zerox even more versatile for users dealing with various document formats.

Thank you for considering this enhancement!

[CarterMcClellan]

Also noted this while constructing a benchmark of the the zerox model against other providers like AWS Textract + Google OCR.

Would be great to have image support!

[pradhyumna85]

duplicate of #67, #58